Initialize model with pretrained words vectors

[Threynaud]

Hi!

I was wondering if it was possible to initialize a GloVe model with one of the pre-trained embeddings before training it on a dataset of my own? I'm not sure to see how it can be done..

Thank you!

[ghost]

Yeah, this is not going to work well due to the optimization setup. But what you can do is train GloVe vectors on your own corpus and then concatenate those with the pretrained GloVe vectors for use in your end application.

[Threynaud]

Thanks for that quick reply! I will try that out!

[niuhuakang]

Hi, I want to know how to "concatenate those with the pretrained GloVe vectors for use in your end application". Is the new trained vectors comparable to the old trained vectors? @Russell91

[JesseTG]

But then wouldn't you lose dependencies between words in the old model and words in the new one?

[sashaostr]

same questions here, does anyone has answers please?

[stevenwernercs]

Am I correct in thinking that GloVe (Global Vectors) is not meant to be appended?
Since it is based on the corpus' overall word co-occurrence statistics from a single corpus known only at initial training time..

[alexandrefelipemuller]

Hi, I couldn't find any solution on all this thread including #67 #77 .. does any one have an source of example?

[ghost]

Bummer

…

On Mon, Nov 4, 2019, 9:16 AM alexandrefelipemuller ***@***.***> wrote: Hi, I couldn't find any solution on all this thread including #67 <#67> #77 <#77> .. does any one have an source of example? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#62?email_source=notifications&email_token=AAIFEMP75T2DUZD6XOCDFETQSBDFRA5CNFSM4DAZFUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEC7ZXXY#issuecomment-549428191>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIFEML7Q55LJV33JPV4TE3QSBDFRANCNFSM4DAZFUQQ> .